The risk of recurrence in women with venous thromboembolism while using estrogens: a prospective cohort study

Transcription

1 Journal of Thrombosis and Haemostasis, 12: DOI: /jth ORIGINAL ARTICLE The risk of recurrence in women with venous thromboembolism while using estrogens: a prospective cohort study L. EISCHER,* S. EICHINGER* and P. A. K Y R L E * *Department of Medicine I, Medical University of Vienna, Vienna; and Karl Landsteiner Institute of Thrombosis Research, Vienna, Austria To cite this article: Eischer L, Eichinger S, Kyrle PA. The risk of recurrence in women with venous thromboembolism while using estrogens: a prospective cohort study. J Thromb Haemost 2014; 12: Summary. Background: The optimal duration of anticoagulation for women who had venous thromboembolism (VTE) associated with estrogen use is unknown. Objectives: To test the hypothesis that women who had a first VTE while using estrogens have a low risk of recurrence. Methods: A Prospective cohort study of 630 women (333 estrogen users, 297 non-users) with a first VTE, who were followed for an average of 69 months after anticoagulation withdrawal. Women with a previous or secondary VTE, coagulation inhibitor deficiency, lupus anticoagulant, cancer, pregnancy, requirement of long-term antithrombotic therapy or homozygosity or double heterozygosity for factor V Leiden and/or the G20210A prothrombin mutation were excluded. The endpoint was objectively documented symptomatic recurrent VTE. Results: VTE recurred in 22 (7%) estrogen users and in 49 (17%) nonusers. After 1, 2 and 5 years, the cumulative probability of recurrence was 1% (95% confidence interval [CI], 0 2), 1% (95% CI, 0 2) and 6% (95% CI, 3 9) among estrogen users and 5% (95% CI, 2 7), 9% (95% CI, 6 13) and 17% (95% CI, 12 22) among non-users. Compared with non-users, estrogen users had an adjusted relative risk (RR) of recurrent VTE of 0.4 (95% CI, ). Compared with non-users in the respective age groups, the RR of recurrence was 0.4 (95% CI, ) among estrogen-containing-contraceptive users and 0.7 (95% CI, ) among women using estrogen-containing menopausal hormone therapy. Conclusions: Women who had their first VTE while using estrogens have a low risk of Correspondence: Paul A. Kyrle, Department of Internal Medicine I, Medical University of Vienna, Waehringer Guertel 18-20, A-1090 Vienna, Austria. Tel.: ; fax: paul.kyrle@meduniwien.ac.at Received 24 October 2013 Manuscript handled by: F. R. Rosendaal Final decision: F. R. Rosendaal, 8 February 2014 recurrent VTE. These women might not benefit from extended anticoagulant therapy. Keywords: estrogens; hormone replacement therapy; pulmonary embolism; recurrence; venous thrombosis. Introduction Venous thromboembolism (VTE) is a chronic disease with the potential of recurrence, especially in patients in whom anticoagulant treatment has been discontinued. The risk of recurrent VTE depends upon the presence (or absence) of certain risk factors. Patients with unprovoked VTE (i.e. those who had VTE in the absence of a transient (temporary) risk factor) are at a particularly high risk of recurrence and are regarded as candidates for extended, even life-long, anticoagulation [1]. For so far unknown reasons, men are at a higher risk of recurrence than women [2,3]. In our analysis of 2004, we reported that the risk of recurrence among women who had their first VTE event during oral contraceptive use was similar to that of women with unprovoked VTE [2]. Findings of other studies regarding the risk of recurrence for female hormone users were controversial. In studies from Canada and France [4,5], the risk of recurrence for oral contraceptive users and non-users was comparable, whereas investigators from the Netherlands, the United States and a multinational patient-level metaanalysis found a lower likelihood of recurrence among women with a hormone-related index thrombosis [3,6,7]. The optimal duration of secondary thromboprophylaxis entails balancing the risk of bleeding during anticoagulant therapy against the risk of recurrent VTE after suspension of anticoagulation. If the risk of recurrence in women using estrogens at the time of VTE is as high as in women with unprovoked VTE, extended anticoagulation might be justified. If these women have a low risk of recurrence, they might not benefit from extended longterm anticoagulation.

2 636 L. Eischer et al The purpose of this report is to provide an update on the risk of recurrence in women with a first VTE associated with estrogen use compared with women in whom the incident VTE was unprovoked. Methods Patients and study design This analysis was performed within the frame of the Austrian Study on Recurrent Venous Thromboembolism (AUREC), an on-going prospective cohort study. Between July 1992 and September 2008 consecutive patients with a first distal and/or proximal deep vein thrombosis (DVT) of the leg and/or pulmonary embolism (PE) who had been treated with anticoagulants for 3 18 months were included. The present analysis includes data collected until March AUREC exclusion criteria were: age younger than 18 years; VTE associated with surgery, trauma, cancer, prolonged immobilization or pregnancy; or requirement for long-term antithrombotic treatment for reasons other than VTE. Patients entered the study at the time of discontinuation of anticoagulation. At study entry, a detailed medical history was obtained. The medical history included a systematic documentation of estrogen use. Two hundred and seventy-five women used estrogen-containing contraceptives (third generation oral contraceptives, 209; first or second generation oral contraceptives, 26; vaginal ring, 2; transdermal patch, 5; unspecified, 33) and 58 women used menopausal hormone therapy (MHT). Of these, 21 women used estrogens alone and 37 estrogens in combination with progesterons. An association between estrogen use and VTE occurrence was assumed when estrogens were used at the time of the VTE. Three weeks after withdrawal of anticoagulation, women were screened for biochemical and genetic risk factors for VTE. Women were excluded when they had deficiency of antithrombin, protein C or protein S, the presence of the lupus anticoagulant, or when they were homozygous or double heterozygous for factor (F) V Leiden and/or the G20210A prothrombin mutation. The women were seen at 6-month intervals for the first year and once a year thereafter, and they were given detailed written information on symptoms of VTE and were asked to report immediately if such symptoms occurred. Female patients were strongly discouraged from further estrogen use. A medical history was obtained at each visit and the use of estrogens was recorded. Patients received thromboprophylaxis with a low-molecular-weight heparin during high-risk situations such as surgery, trauma, prolonged immobilization or long-haul air travel. The ethics committee of the Medical University of Vienna approved the study and all patients gave written informed consent prior to inclusion in the study. Diagnosis of VTE As previously described, the diagnosis of VTE was established by a positive finding on venography, colour duplex sonography, ventilation-perfusion lung scanning or spiral computed tomography (CT) [8]. Study endpoints The endpoint of the study was recurrent symptomatic DVT (including distal DVT) confirmed by venography or colour duplex sonography, or recurrent symptomatic PE confirmed by multi-slice CT, ventilation-perfusion lung scanning or autopsy. Recurrent DVT was diagnosed if the patient had an abnormal compression ultrasound or an intraluminal filling defect on venography in the case of documented recanalization of the initial thrombus, a thrombus in another deep vein in the extremity involved in the previous event, a thrombus in the opposite extremity, or a thrombus in the same venous system with a proximal extension of the thrombus (if the upper limit of the original thrombus had been visible) or the presence of a constant filling defect surrounded by contrast medium (if the original thrombus had not been visible). Laboratory analysis Venous blood from fasting patients was collected into EDTA tubes or in 1 : 10 volume of 0.11 mm trisodium citrate and immediately centrifuged for 20 min at g. Screening for FV Leiden and factor II G20210A, and measurement of antithrombin, protein C, protein S and the lupus anticoagulant, were carried out by standard methods. Statistical analysis Categorical data were compared among groups using contingency-table analyses (the chi-square test). Continuous data (presented as means SD) were compared by means of Mann Whitney U-tests. Survival-time methods were used to analyze the time to recurrence among patients with a subsequent episode of VTE (uncensored observations) or the duration of follow-up among patients without recurrence (censored observations). Data on patients who left the study were censored at the time of withdrawal and data on patients who were lost to follow-up were censored at the time of their last visit. The probability of recurrence was estimated according to the method of Kaplan and Meier. To test for homogeneity between strata, we applied the log-rank test. Univariate and multivariate Cox proportional-hazards models were used to analyze the association between estrogen use and the risk of recurrent VTE. Analyses were adjusted for age, presence or absence of FV Leiden and site of VTE.

3 Recurrence risk after estrogen-related VTE 637 Data are given as mean (SD) unless otherwise indicated. P values were two-tailed, and were considered as indicating statistical significance if lower than SPSS software (version 20; IBM, Armonk, NY, USA) was used. Results Study population The total study population consisted of 630 women who were on average years of age and had been treated with oral anticoagulants for 7 3 months. The location of the incident VTE was DVT in 361 (57%) and PE DVT in 269 (43%) women. One hundred and fiftyfour (24%) women were heterozygous carriers of the FV Leiden mutation. VTE had occurred in association with estrogen use in 333 women (53%). The duration of follow-up was months. Ninety-five women left the study because they required antithrombotic treatment for causes other than venous thrombosis, or because they were given a diagnosis of cancer (15), because they became pregnant (62) or for other reasons (12). Forty-one women were lost to followup. Twelve women died during follow-up for reasons other than recurrent VTE: four from cardiac failure, two from cancer, three had an ischemic stroke, two died from intracranial bleeding and one committed suicide. None of the women used estrogens after withdrawal of anticoagulation. Recurrence of VTE Recurrent VTE was recorded in 71 of the 630 female patients (11%). Thirty-six women had DVT only and 35 women had PE with or without DVT. PE was fatal in two women. The probability of recurrence was 3% (95% CI, 1 4), 5% (95% CI, 3 7) and 11% (95% CI, 8 14) after 1, 2 and 5 years, respectively. Recurrence of VTE according to estrogen use Table 1 shows the characteristics of women with unprovoked VTE in comparison to women who had their VTE while using estrogens. Women who used estrogens were younger than non-users and their observation time was longer. The proportion of women carrying FV Leiden was higher among estrogen users than among non-users. Recurrent VTE occurred in 22 of 333 estrogen users (7%) and in 49 of 297 non-users (17%). After 1, 2 and 5 years, respectively, the probability of recurrence was 1% (95% CI, 0 2), 1% (95% CI, 0 2) and 6% (95% CI, 3 9) among estrogen users and was 5% (95% CI, 2 7), 9% (95% CI, 6 13) and 17% (95% CI, 12 22) among non-users (Fig. 1). Compared with non-users, estrogen users had a relative risk (RR) of recurrent VTE of 0.3 (95% CI, ) before and 0.4 (95% CI, ) after Table 1 Patient characteristics according to estrogen use Non-users (n = 297) Users (n = 333) P Age (years) < Site of first VTE; n (%) Distal DVT 69 (23) 78 (23) 0.7 Proximal DVT 105 (36) 109 (33) PE DVT 123 (41) 146 (44) Duration of anticoagulation (months) Observation < time (months) Factor V Leiden; n (%) 48 (16%) 98 (28%) < Plus-minus values are mean SD. Cumulative probability of recurrence (%) P < Years after discontinuation of anticoagulation Number at risk No estrogen use Estrogen use adjustment for age, site of VTE and FV Leiden carrier status (Table 2). The low risk of recurrence among estrogen users was confined to women who had their first VTE event in association with estrogen-containing-contraceptive use. After 5 years, the probability of recurrence was 4% (95% CI, 1 7) among estrogen-containing-contraceptive users and was 17% (95% CI, 10 24) among non-users. Compared with non-users from the same age group (18 and 55 years), the RR of recurrence among estrogen-containing-contraceptive users was 0.3 (95% CI, ) before and 0.4 (95% CI, ) after adjustment (Table 2). After 5 years, the probability of recurrence was 14% (95% CI, 3 24) among MHT users and was 18% (95% CI, 12 24) among non-users. Compared with non-users from the same age group (45 and 77 years), both the crude and the adjusted RR of recurrence among women using estrogen-containing menopausal hormone therapy was 0.7 (95% CI, ) (Table 2) No estrogen use Estrogen use Fig. 1. Cumulative probability of recurrent venous thromboembolism (VTE) among women with unprovoked first VTE (solid line) and women with a first VTE while using estrogens (dotted line)

4 638 L. Eischer et al Table 2 Recurrence risk according to estrogen use categories Discussion Events/patients (n/n) Unadjusted RR Adjusted RR (95% CI)* No estrogens 49/ Estrogens 22/ ( ) 0.4 ( ) No estrogen-containing 27/ contraceptives Estrogen-containing 14/ ( ) 0.4 ( ) contraceptives No menopausal 39/ hormone therapy Menopausal hormone therapy 8/ ( ) 0.7 ( ) *For age, site of venous thromboembolism (distal deep vein thrombosis (DVT), proximal DVT, pulmonary embolism) and factor V Leiden. The principal finding of our study is that the risk of recurrent VTE is 60% lower among women who had their first VTE while using estrogens as compared to women with unprovoked VTE. Five years after withdrawal of anticoagulation, the likelihood of recurrence was 17% among estrogen non-users and only 6% among estrogen users (with an upper 95% confidence bound of 9%). Of note, the low recurrence risk among estrogen users was found in younger women taking estrogen-containing contraceptives rather than in older women receiving MHT. So far, data on the recurrence risk among women who developed their first VTE while using estrogens are scarce and controversial. In a sub-analysis of REVERSE, a large prospective cohort study from Canada, the likelihood of recurrent VTE was studied in 67 estrogen users (49 oral contraceptive users and 18 MHT users) and in 259 nonusers. All women had a first unprovoked proximal DVT and/or PE. Oral contraceptive users had a lower annual risk of recurrence than non-users (1.7; 95% CI, vs. 5.0; 95% CI, ). With an adjusted hazard ratio (HR) of 0.6 (95% CI, ), this difference, however, did not reach statistical significance. Interestingly, the authors reported a 2-fold higher risk of recurrence in MHT users than in non-users [4]. In a prospective cohort study from France, 180 women with VTE while exposed to combined oral contraception and 61 women with thrombosis not exposed to female hormones were followed for an average of 66 months. Only 19 recurrences were observed, 14 in estrogen users and five in non-users. After adjustment for age, no association was found between exposure to estrogens and the incidence rates of recurrence for an incidence ratio of 0.7 (95% CI, ) [5]. Cushman and co-workers prospectively studied 109 women without unprovoked hormone-related incident VTE and 123 women with unprovoked VTE associated with oral contraceptives (n = 42) or MHT (n = 81). Women with hormone-related thrombosis had a 46% lower risk of recurrence than other women (HR, 0.54; 95% CI, ) [7]. In the Leiden Thrombophilia Study, 70 women used oral contraceptives at the time of their first thrombotic event and refrained from estrogen use thereafter. The recurrence rate in women who had never used oral contraceptives was slightly, albeit not statistically significantly, higher than in women who used oral contraceptives at the time of VTE (16.2 per 1000-patient years; 95% CI, per 1000 patient-years; vs. 9.7 per 1000-patient years; 95% CI, )[6]. In a patient-level meta-analysis of seven prospective clinical studies (including AUREC), the risk of recurrent VTE among estrogen users was 50% lower as compared with that of non-users (HR, 0.5; 95% CI, ). However, in this analysis the mean follow-up of the total patient population was shorter than in our study and the type of hormone used was specified in < 20% of women [3]. Is there an explanation for our finding of a low risk of recurrent VTE among estrogen-containing-contraceptive users? Estrogens induce a hypercoagulable state and, most likely by interacting with other genetic and/or acquired risk factors, facilitate VTE development. The risk of thrombosis subsides when women avoid estrogens after stopping anticoagulation. Women in whom VTE develops in the absence of estrogens (or another temporary risk factor) may have a lower VTE threshold due to other (unknown) reasons. The finding of a low recurrence risk among women who developed VTE while using estrogens has important clinical implications, particularly for estrogen-containingcontraceptive users. The optimal duration of anticoagulation entails balancing the risk of bleeding against the risk of recurrent VTE after stopping anticoagulant therapy. The rate of major bleeding during vitamin K antagonist therapy in VTE patients is approximately 3% during the first 3 months. Data on the bleeding risk thereafter are scarce. Depending on the risk profile, the estimated annual risk of major bleeding ranges between 0.8% and more than 6% [1]. Outside clinical studies in unselected patients receiving usual care the bleeding risk associated with vitamin K antagonist treatment is most probably higher [9]. With a likelihood of recurrent VTE of < 10% 5 years after stopping anticoagulation, we strongly believe that women with a first VTE related to estrogen use (and in particular women who had VTE while using estrogencontaining contraceptives) will not benefit from extended thromboprophylaxis and should receive anticoagulants for a very limited period of time. The proportion of carriers of the FV Leiden mutation was almost 2-fold higher in women who developed VTE while using estrogens than in women who did not. VTE is a multi-causal disease and there are numerous examples of risk factor interactions causing a boost in the risk of

5 Recurrence risk after estrogen-related VTE 639 thrombosis. For example, Vandenbroucke and co-workers reported in 1994, that both premenopausal women using estrogen and female carriers of the FV Leiden mutation have a higher risk of a first VTE than non-users or women without FV Leiden (4-fold and 8-fold, respectively). By far the highest risk (30-fold), however, was found in estrogen-using women carrying the FV Leiden mutation [10]. The relatively large number of estrogen users with FV Leiden in our study can thus be readily explained by a selection effect resulting from risk factor interaction. This imbalance in FV Leiden distribution between the two groups, we believe, did not affect the outcome of our analysis as (i) all women refrained from further estrogen use (i.e. this risk factor was no longer present) and (ii) the impact of FV Leiden on the risk of recurrence is minimal both in our cohort as well as in other studies [11,12]. The risk of recurrence among women using MHT at the time of their incident VTE was not substantially lower than that of estrogen non-users. This observation implies that the duration of anticoagulant treatment should not differ for older women with MHT-associated VTE and women with unprovoked VTE. However, the number of MHT-using women in our study was not large enough to justify any such clinical management decisions. Our study has strengths and limitations. With more than 600 women, of whom almost 350 had an incident VTE while using estrogen-containing hormones, this is the largest single-center study so far. Women were prospectively followed on average over more than 5 years and were seen at our institution at regular intervals. None of the women, regardless of whether they had their VTE while using estrogens or not, refrained from estrogen intake during follow-up. Losses to follow-up were low (6.5%). The study population was homogenous: we only studied women with a first unprovoked VTE or with VTE linked to estrogen use, whereas women who had VTE related to other precipitating temporary risk factors were excluded. Of note, exclusion of women with an unprovoked distal DVT without PE from the analysis did not affect the results (data not shown). Our data cannot be applied to some other patient groups. Women with a natural coagulation inhibitor deficiency or a lupus anticoagulant or when they were homozygous or double heterozygous for FV Leiden and/or the G20210A prothrombin mutation were regarded as candidates for indefinite anticoagulant therapy and, hence, excluded from the study. Although we had information on the type of hormone use, the patient numbers in the respective groups were not large enough to perform comparative analyses. The number of MHT users was small, precluding counseling these women with respect to the optimal duration of treatment. Finally, baseline imaging at the completion of therapy was not systematically performed, which potentially could lead to over-diagnosis of recurrent VTE. In conclusion, we found a low risk of recurrent VTE among women who had their first VTE while using estrogen-containing hormones. The recurrence risk was particularly low in estrogen-containing-contraceptive users. These women might not benefit form extended anticoagulation. Further clinical studies are, however, warranted. Addendum P. A. Kyrle had full access to all data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. P. A. Kyrle and S. Eichinger: study concept and design. L. Eischer: acquisition of data. L. Eischer, S. Eichinger and P. A. Kyrle: analysis and interpretation of data. P. A. Kyrle: drafting of the manuscript. L. Eischer and S. Eichinger: critical revision of the manuscript for important intellectual content. L. Eischer: statistical analysis. P. A. Kyrle and S. Eichinger: obtained funding. P. A. Kyrle: study supervisor. Acknowlegdements This work was supported by grants from the Medizinischwissenschaftlicher Fonds des B urgermeisters der Bundeshauptstadt Wien and the Jubil aumsfonds of the Austrian National Bank. This work was presented as an oral communication at the 23rd Congress of the International Society of Thrombosis and Haemostasis in Amsterdam on 1 July Role of the sponsor The sponsor had no involvement in: the design and conduct of the study; the collection, management, analysis and interpretation of the data; or the preparation of the manuscript. Disclosure of Conflict of Interests The authors state that they have no conflict of interests. References 1 Kearon C, Akl EA, Comerota AJ, Prandoni P, Bounameaux H, Goldhaber SZ, Nelson ME, Wells PS, Gould MK, Dentali F, Crowther M, Kahn SR, American College of Chest Physicians. Antithrombotic therapy for venous thromboembolic disease. Chest 2012; 141: e419s 94S. 2 Kyrle PA, Minar E, Bialonczyk C, Hirschl M, Weltermann A, Eichinger S. The risk of recurrent venous thromboembolism in ten and women. N Engl J Med 2004; 350: Douketis J, Tosetto A, Marcucci M, Baglin T, Cosmi B, Cushman M, Kyrle P, Poli D, Tait RC, Iorio A. Risk of recurrence after venous thromboembolism in men and women: patient level meta-analysis. BMJ 2011; 342: d813. doi: /bmj.d Le Gal G, Kovacs MJ, Carrier M, Do K, Kahn SR, Wells PS, Anderson DA, Chagnon I, Solymoss S, Crowther M, Righini M, Lacut K, White RH, Vickars L, Rodger M. Risk of recurrent

6 640 L. Eischer et al venous thromboembolism after a first oestrogen-associated episode. Data from the REVERSE cohort study. Thromb Haemost 2010; 104: Le Moigne E, Delluc A, Tromeur C, Nowak E, Mottier D, Lacut K, Le Gal G. Risk of recurrent venous thromboembolism among young women after a first event while exposed to combined oral contraception versus not exposed to: a cohort study. Thromb Res 2013; 132: Christiansen SC, Cannegieter SC, Koster T, Vandenbroucke JP, Rosendaal FR. Thrombophilia, clinical factors, and recurrent venous thrombotic events. JAMA 2005; 293: Cushman M, Glynn RJ, Goldhaber SZ, Moll S, Bauer KA, Deitcher S, Shrivastava S, Ridker PM. Hormonal factors and risk of recurrent venous thrombosis: the prevention of recurrent venous thromboembolism trial. J Thromb Haemost 2006; 4: Eichinger S, Heinze G, Jandeck LM, Kyrle PA. Risk assessment of recurrence in patients with unprovoked deep vein thrombosis or pulmonary embolism: The Vienna prediction model. Circulation 2010; 121: Beyth RJ, Quinn L, Landefeld CS. A multicomponent intervention to prevent major bleeding complications in older patients receiving warfarin. A randomized, controlled trial. Ann Intern Med 2000; 133: Vandenbroucke JP, Koster T, Briet E, Reitsma PH, Bertina RM, Rosendaal FR. Increased risk of venous thrombosis in oral-contraceptive users who are carriers of factor V Leiden mutation. Lancet 1994; 344: Eichinger S, Pabinger I, St umpflen A, Hirschl M, Bialonczyk C, Schneider B, Melichart M, Rintelen C, Speiser W, Mannhalter C, Minar E, Lechner K, Kyrle PA. The risk of recurrent venous thromboembolism in patients without and with factor V Leiden. Thromb Haemost 1997; 77: Kyrle PA, Rosendaal FR, Eichinger S. Risk assessment for recurrent venous thrombosis. Lancet 2010; 376: